Model-based Design of Hybrid Electrical Vehicle Devices using automated Optimization Strategies

Only the simulation of complete models allows the user to gain detailed insight into the often highly non-linear model behavior. The interactions between driver, combustion engine, electric motors, battery, powertrain, chassis, and road conditions are important factors in finding the optimal parameter set. The model based design of a complete HEV requires extremely fast models when optimization strategies are applied, as they often require several hundred calls of the model to determine an optimal parameter set. This demand for high model speed causes many simulation tools to fail. In this paper, we demonstrate the integration of an extremely fast high-fidelity seriesHEV model with advanced optimization strategies. Besides determining an optimal parameter set, the analysis of sensitivities and the robustness while satisfying the model constraints are considered.